Robust Adaptive Control For Formations Of Marine Surface Vessels

The dynamics of marine surface vessel is characterized by intrinsic nonlinear, nonholonomic constraint, underactuation and time-varying disturbances. Thus, the motion control problem of marine surface vessel received significant attention. The marine control community focused on the motion control problem of single vessel. However, with the development of marine technology, there is increasingly need for utilizing multiple vessels to perform difficult tasks that cannot be performed by a single one. The underlying applications include marine survey operations, environment monitoring, underway replenishment, cooperative search and rescue, and so on. For these applications, the formation control technique is undoubtedly promising. The marine control community has focused on the pursuit of effective methods to achieve the formation control, and many good results have been obtained; however, there are still problems that need to be settled. The consistent explores of formation control techniques to achieve coordinated control of marine vessels are of great significance.This thesis is concerned with the formation control problem of surface vessels with uncertain dynamics. The main contributions of this thesis are as follows:First, we considered the leader-follower formation control problem of surface vessel with uncertain leader dynamics and uncertain follower dynamics. Within the leader-follower formation control framework, by online approximating the uncertain dynamics, a leader-follower robust adaptive formation control design is developed for surface vessel using backstepping design technique. Based on the Lyapunov stability analysis, it is proved that all signals of closed-loop system are semiglobally uniformly ultimately bounded (SUUB). The numerical simulations verified the effectiveness of control algorithm and learning ability of neural network.Second, to deal with the complexity problem of using backstepping technique, a dynamic surface control design approach is introduced. By online approximating the uncertain dynamics, a robust adaptive neural network formation controller is developed. Based on Lyapunov stability analysis, it is proved that the formation tracking errors converge to a small neighborhood of origin. Comparative studies with traditional control method are performed to show the effectiveness of the proposed control method. Third, we considered the formation tracking problem of underactuated surface vessel with unknown nonlinear dynamics and unknown disturbances from the environment. Based on the pure-pursuit guidance principle of missiles and point-to-point navigation of vessels, at first, a formation control method is proposed using target tracking mechanism; then, by utilizing the backstepping technique and neural network approximation, a robust adaptive leader-follower formation control scheme is developed for surface vessel. Based on Lyapunov stability analysis, it is proved that the target tracking errors converge to a small neighborhood of origin. Simulation results demonstrated the effectiveness of proposed method.Forth, the leader-follower motion synchronization problem of surface vessel with uncertain dynamics is considered under undirected network with fixed communication structure. By incorporating the dynamic surface design approach, neural network and leader-follower synchronization control strategy, a distributed robust adaptive formation control method is proposed. At first, the case where the followers have full access to the leader is considered and a general formation control algorithm is proposed; and then, the above result is extended to two cases:one considers the communication delays and the other only partial followers have access to the leader. Based on Lyapunov stability analysis, it is proved that all signals of the closed-loop system are SUUB and cooperative tracking errors converge to a small neighborhood of origin. Extensive studied are given to show the effectiveness of the method.Finally, under directed and fixed communication structure, the flocking problem of nonholonomic multi-agent systems with uncertain dynamics is considered. Besides, the leader motion to be synchronized is also to be unknown. By online approximating the uncertain follower dynamics and using a bounded artificial potential function to avoid the collisions between the agents, a neural flocking algorithm is proposed. Under the assumption that the network is strong connected, it is proved that the agents' velocities and headings are synchronous with the leader, and collisions between agents can be avoided. Simulation results verified the effectiveness of the neural flocking control algorithm.